Introduction

• Know if a server is sized correctly in terms of processing power;
• Determine times of day when CPU usage level is high

Determine how many processes are contending for CPU resources simultaneously

Know which processes are causing a CPU spike on the server

Know which servers have high CPU utilization, and which ones are under-utilized

• Track free memory availability on your servers;
• Determine if your servers are adequately sized in terms of memory availability

Determine servers with high swap usage

Know which processes are taking up memory on a server

Know which servers have the lowest free memory available and hence, may be candidates for memory upgrades

• Track the number of processes blocked on I/O;
• Indicates if there is an I/O bottleneck on the server

• Track the percentage of time that the disks on a server are heavily used.
• Compare the relative busy times of the disks on a server to know if you can better balance the load across the disks of a server

Monitor disk read and write times to detect instances when a disk is slowing down (Windows only)

Track the number of processes queued on each disk drive to determine disk drives that may be responsible for slow downs

Determine which processes are causing disk reads/writes

• Determine how long a server has been up;
• Track times when a server was rebooted;
• Determine times when unplanned reboots happened;

Know which servers have not been rebooted for a long time;

Know the total capacity of each of the disk partitions of a server

• Track the free space on each of the disk partitions of a server;
• Proactively be alerted of high disk space levels on a server;

Monitor and alert on page file usage of a Windows server;

• Track the bandwidth usage of each of the network interfaces of a server (Windows only);
• Identify network interfaces that have excessive usage

• Determine queuing on each of the network interfaces of a server;
• Identify network interfaces that may be causing a slowdown;

• Track the traffic into and out of a server through each interface;
• Identify servers and network interfaces with maximum  traffic;

• Track the quality of a network connection to a server;
• Identify times when excessive packet loss happens;

Determine the average delay of packets to a server;

Determine times when a server is not reachable over the network;

Track currently established  TCP connections to a server;

Monitor the server workload by tracking the rate of TCP connections to and from a server

• Track the percentage of TCP segments retransmitted from the server to clients;
• Be alerted when TCP retransmits are high and therefore, are likely to cause significant slow*downs in application performance;

• Track the number of processes of a specific application that are running simultaneously;
• Identify times when a specific application process is not running

• Monitor the CPU usage of an application over time;
• Determine times when an application is taking excessive CPU resources.

• Track the memory usage of an application over time;
• Identify if an application has a memory leak or not;

Track the number of threads running for an application’s process (Windows only);

• Track the number of handles held by an application over time (Windows only);
• Identify if a process has handle leaks;

Determine if a service is running or not

• Track the number of information, warning, and error events logged in the Microsoft Windows System and Application event logs;
• Correlate events in the Windows event logs with other activity on the server (e.g., service failure)
• Obtain details of the events in the event logs;

• Monitor all events logged in the Microsoft Windows Security log;
• Obtain details of all failure events;

Track and be alerted of all errors logged in the /var/adm/messages log of a Unix system